01:43 | _dab_ | brcm4330 includes on-chip 2.4GHz and 5 GHz WLAN. Whats the story on the 5GHz band? |
03:48 | msalerno | _dab_ Last I checked there was no 5Ghz |
03:49 | _dab_ | Is that hardware or software. The broadcom site says brcm4330 has 5GHZ |
03:50 | _dab_ | http://www.broadcom.com/products/Wireless-LAN/802.11-Wireless-LAN-Solutions/BCM4330 |
03:55 | msalerno | Let me find the post - 1 sec |
03:59 | msalerno | I had a spec sheet from a forum post long ago. I can't find it |
03:59 | msalerno | it basically said 5ghz was not part of the hw |
04:02 | msalerno | http://www.solid-run.com/community/topic961.html |
04:34 | _dab_ | thanks msalerno |
06:54 | sanjeevs | Hello Folks |
06:55 | sanjeevs | I am working on upgrading kernel from 3.0.35 to 3.14 and found that neither freescale nor Boundry devices has achived that |
06:56 | sanjeevs | But I found that Solid Run is currently working on 3.14 kernel targeting CuBox-i and HummingBoard. |
07:08 | sanjeevs | I would like to know the way how fsl patches/bugfixes has been tracked and ported on 3.14 mainline tree |
07:10 | sanjeevs | is anyone can let me know which is the fastest way/technique to achive this |
07:11 | Humpelst1lzchen | sanjeevs: xbmc and gstreamer have hw accel on 3.14 |
07:15 | sanjeevs | Hello Humpelst1lzchen: I have 3.14.19 stable kernel and my intention is to port the Freescale bsp patches on mainline kernel.I could see that branch linux-linaro-lsk-v3.14-mx6 contain's fixes from freescale bsp release and other contributors.I wanted to know from where I get the only freescale bsp patches and in which order they need to apply in my mainline tree |
07:19 | sanjeevs | Does anyone has any idea how linux-linaro-lsk-v3.14-mx6 getting tracked ? |
08:23 | jnettlet | sanjeevs, all the tracking is done by me. I track both upstream submissions for drivers as well as fsl's git repository and merge the patchsets that seem most usable and stable. |
08:25 | jnettlet | generally I only track iMX6s/dl/d/q patches as that is all the FSL hardware that SR currently makes. |
08:31 | sanjeevs | Thanks jnettlet I saw your name here http://www.solid-run.com/community/topic1525.html |
08:32 | jnettlet | I am currently working on porting the latest driver set over to the latest 3.14.30 kernel |
08:32 | sanjeevs | So you are doing cherry-picking of commit sequentially from Tree http://git.freescale.com/git/cgit.cgi/imx/linux-2.6-imx.git or merging any branch i.e specific to MX6s/dl/d/q patches maintain by Freescale |
08:34 | jnettlet | I review their patchset vs what they have pushed upstream favoring the upstream patches. Then I merge any additional patchsets that fix problems or add functionality that seems valid. |
08:35 | sanjeevs | Another query is for example you have taken an base version of mainline kernel i.e 3.14.14 and then how you identifying commit which need to cherry-pick ? Are you doing on basis of kernel subystem or taking diff of mainline & Freescale kernel |
08:35 | jnettlet | I try to merge patchsets by driver instead of all mixed in so they are easier to debug driver problems with bisects |
08:36 | sanjeevs | yup make sense and Freescale Soc & DT you track different correct |
08:36 | jnettlet | that is where things git a bit tricky. I try to track mainline but in some cases it is just to difficult so I adapt to FSL's DT |
08:36 | sanjeevs | because i.e tested and proven correct |
08:37 | sanjeevs | Do we know how we identify what patches from Freescale BSP has been upstreamed ? |
08:37 | jnettlet | well my ultimate hope is as time goes on and more FSL patches are merged upstream, I will be able to have more mainline patches and less BSP patches |
08:38 | sanjeevs | But there are large no of patches around 2202 pacthes above 3.14.14 |
08:38 | jnettlet | I have a script that tries to best match BSP patches vs upstream commits, but it is more a general approach. |
08:39 | jnettlet | yes, about half of those are upstream, or pending upstream, and the other half is BSP. |
08:39 | jnettlet | with the first 3.10 kernel it was more like 20/80 |
08:39 | sanjeevs | and then rest of the commit's you cherry-pick one by one bases on Fsl Bsp release version |
08:40 | jnettlet | I review the patches in bulk and then have scripts that merge my approved patchsets. |
08:40 | sanjeevs | and the sequence is same how they are applied on fsl-imx tree correct ? |
08:40 | jnettlet | then of course there are patches I need to massage to apply to the upstream base |
08:42 | jnettlet | the sequence per driver is the same. |
08:43 | jnettlet | and then I carry a hundred or so patches of my own fixes, community fixes, and also work boundary has done. |
08:43 | sanjeevs | so you do driver by driver or in otherwards you have taken for example audio in driver and then you cherry-pick from freescale-imx kernel and applied all the change on top of mainline kernel |
08:45 | sanjeevs | I am planning to do the same and I would like to divide first into various parts like soc fixes,Driver,DT ,GPU, VPU and IPU so on |
08:46 | sanjeevs | I came across shawnguo trre https://git.kernel.org/cgit/linux/kernel/git/shawnguo/linux.git/?h=imx/soc which also maintain the freescale changes |
08:47 | jnettlet | shawnguo worked for FSL but I believe is now working at Linaro. |
08:47 | jnettlet | but he is one of the upstream maintainers |
08:48 | jnettlet | sanjeevs, is there any reason that you just don't want to work off my kernel and add support for whatever you are working on? |
08:48 | jnettlet | we are working on merging patches for additional boards like the wandboard, and utilite |
08:49 | jnettlet | I am going to merge those after I finish working out the latest BSP release |
08:51 | sanjeevs | jnettlet,yes of course I can fork your kernel tree. |
08:52 | sanjeevs | I would like to support the Nitrogen6x Board which is very similar to the sabrelite Board |
08:53 | jnettlet | sanjeevs, sure. please send pull requests for any patches. I will be glad to bring in support for that series of boards. |
08:53 | sanjeevs | Your main focus is on cubeBox-i and HummingBoard I beliver |
08:53 | sanjeevs | I can send you to pull patches from boundry device how to add support for these board |
08:54 | sanjeevs | you work with solidrun I believe |
08:55 | sanjeevs | Patcheset you referring in our discussion is pacthees/fixes per BSP http://git.freescale.com/git/cgit.cgi/imx/linux-2.6-imx.git/log/?h=imx_3.10.17_1.0.1_ga |
08:56 | sanjeevs | Jnettlet:I believe you are doing the similar work shawnguo doing and why you didn't cloned his tree |
08:59 | jnettlet | shawnguo is working on upstreaming patches to mainline. I am working on backporting mainline patches to the latest Linux LTS tree. |
09:00 | jnettlet | maintaining BSP patches on top of mainline is too time consuming because of the constant kernel churn. That is why I rely on LTS to provide a most stable base that will still get security/critical updates for 2 years. |
09:01 | sanjeevs | yes i.e correct |
09:01 | jnettlet | I chose Linaro's LTS to base off of because then it is easier to merge the patchsets over to an Android branch they maintain, or the -rt patchset branch they maintain |
09:02 | sanjeevs | IMO same could be achived over stable mainline kernel |
09:02 | sanjeevs | I have considered the base from stable mainline tree |
09:04 | sanjeevs | maybe you have found Linaro LTS more stable then Mainline stable tree |
09:06 | sanjeevs | Till when you are availble on IRC ? I will have lunch quickly and get back to you |
09:06 | jnettlet | I generally here most the day |
09:07 | sanjeevs | OK ! |
09:07 | sanjeevs | Do you work for SolidRun ? |
09:08 | sanjeevs | Did you describe the maintaning process somewhere |
09:09 | sanjeevs | Patcheset you referring in our discussion is pacthees/fixes per BSP http://git.freescale.com/git/cgit.cgi/imx/linux-2.6-imx.git/log/?h=imx_3.10.17_1.0.1_ga |
09:36 | jnettlet | sanjeevs, yes I work with SolidRun. I do not have any documentation on the process because really nobody has asked before. |
09:37 | jnettlet | yes that is the FSL repository for their BSP kernel |
10:41 | sanjeevs | Thanks Jnettlet |
10:46 | sanjeevs | Here is the summary based on our Discussion: 1) Clone upstream stable kernel 3.14 from mainline or Linaro Fsl stable kernel 2) Divide the Task into different subsystem 3)cherry-pick the change from Fsl patchset starting from imx_3.5.7_1.0.0_alpha imx_3.10.17_1.0.0_beta,imx_3.10.17_1.0.0_ga can be found here linux-2.6-imx.git - Freescale i.MX Linux Tree to mainline kernel version selected 4)review the patch set before merging 5) resolve the conflicts if |
11:18 | sanjeevs | can you have a look |
11:38 | rabeeh | sanjeevs: jnettlet is out now; but i think he would nod for the summary |
12:32 | sanjeevs | Thanks rabeeh |
22:10 | Artox | does anyone know where to get libfslaudiocodec as used by gstreamer-imx-1.0? |
22:10 | Artox | dv__? |
22:12 | mk01 | yocto for instance |
22:12 | Artox | I mean |
22:13 | Artox | tried looking here but didnt find anything: https://github.com/Freescale/meta-fsl-arm |
22:13 | Artox | I mean the raw package as provided by freescale |
22:13 | Artox | or is it just part of libfslcodec |
22:14 | mk01 | wait |
22:14 | Artox | it seems gstreamer-imx configure script just looks for libfslaudiocodec.pc |
22:14 | mk01 | yes |
22:14 | mk01 | there are two |
22:14 | mk01 | code |
22:14 | mk01 | codec |
22:14 | mk01 | and aaccodec |
22:14 | mk01 | codec is all audio codecs but aac |
22:14 | mk01 | which is distributed separate |
22:16 | Artox | can you tell me if tehre is such a .pc file? |
22:16 | Artox | in yocto? |
22:24 | Artox | hah |
22:24 | Artox | it *is* part of libfslcodec |
22:24 | Artox | I jsut forgot to package the damn .pc |
22:26 | mk01 | Artox: |
22:26 | Artox | ? |
22:26 | mk01 | on apt.xbian.org / staging repo |
22:26 | mk01 | imx6-jessie |
22:26 | Artox | jessie is not stable |
22:26 | mk01 | there are those packages |
22:26 | Artox | and, yes |
22:26 | Artox | but probably not done by dpkg-buildpackage |
22:27 | Artox | also, I packaged that stuff months ago |
22:27 | Artox | all on github |
22:27 | Artox | I just need to add the .pc |
22:27 | Artox | when I started I didnt realize xbian was there for what I wanted |
22:27 | mk01 | I don't remember even if it has .pc |
22:27 | Artox | so I made my own debian packages |
22:27 | mk01 | nothing uses those libs |
22:27 | Artox | still not all finished, sadly |
22:28 | mk01 | there is one nore |
22:28 | mk01 | more |
22:28 | mk01 | fslparser |
22:28 | Artox | and that provides the aac? |
22:28 | mk01 | probably mp3 header parsing |
22:28 | Artox | ah |
22:28 | Artox | well, now gstreamer-imx does |
22:28 | mk01 | prefix=/usr/local |
22:28 | mk01 | exec_prefix=${prefix} |
22:28 | mk01 | libdir=${exec_prefix}/lib/imx-mm/parser |
22:28 | Artox | and I intend to get it |
22:28 | mk01 | includedir=${prefix}/include/imx-mm/parser |
22:28 | mk01 | Name: Freescale media format parser |
22:29 | mk01 | Description: Optimized SW media format parser from freescale |
22:29 | mk01 | Version: 3.0.8 |
22:29 | mk01 | Libs: -L${libdir} -L${exec_prefix}/lib |
22:29 | mk01 | Cflags: -I${includedir} |
22:29 | mk01 | it is one the apt too (If you would need it) |
23:06 | Artox | umm |
23:06 | Artox | libfslcodec is softfloat only? |
23:06 | Artox | is that possible? |
23:06 | Artox | and is that a problem? |
23:06 | Artox | oh |
23:06 | Artox | pass --enable-fhw to configure |
23:06 | Artox | weird |
23:12 | mk01 | Artox: that pkgs (files) were distributed in that (binary) form only |
23:12 | Artox | yes |
23:12 | Artox | my problem was ubuntus dpkg-shlibdeps |
23:12 | Artox | it was going real mad |
23:12 | Artox | so I just disabled it |
23:13 | Artox | debian wheezy's didnt seem to care |
23:13 | mk01 | that's probably the v2 vs. v5 requirements |
23:14 | mk01 | (of dpkg-XXX tools) |
23:14 | Artox | ah |
23:14 | Artox | well, now time to write a spec file |
23:17 | Artox | mk01, if you got any questionsusing the obs, automagically uploading sources, etc feel free to ask me or the TrojanHorse |
23:17 | mk01 | malte just told me |
23:17 | mk01 | will do |
23:22 | mk01 | Artox: how to select a branch on git remote URL ? |
23:24 | Artox | ah |
23:24 | Artox | in the _service file? |
23:24 | Artox | you just set revision to the branch name |
23:25 | Artox | works just fine |
23:26 | mk01 | ok |